The Pentagon's Brain (8 page)

Teller told Starbird that the test ban was a threat to national security. That it showed weakness and vulnerability and opened America up to a sneak nuclear attack. “The purpose [of this telegram] is in part to clarify laboratory plans and in part to point out dangers in connections with future discussions concerning the test moratorium,” Teller wrote. “The laboratory must continue research and development of nuclear weapons,” he wrote, “in order to comply with the [president’s] directive.” More tests needed to be done in order to make sure that it was safe to comply. Furthermore, he argued, many of Livermore’s nuclear tests were not tests per se but rather scientific experiments, as Operation Argus was.

There was a loophole to be explored, Teller suggested. “Explosions below a kiloton cannot be detected and identified by any of the methods considered realistic by any of the delegations at the Geneva Conference,” he wrote. The United States could secretly conduct low-yield tests. Yes, it would be cheating, but the Russians could not be trusted, and surely they would cheat too.

E
ugene McManus, an electronics technician, worked at the top of the world. He had joined the Air Force, at age seventeen, for adventure and to learn radar technology, and now here he was four years later working at a classified ARPA-activated outpost just nine hundred miles from the North Pole. This was the Ballistic Missile Early Warning System (BMEWS) facility, the world’s first operational missile-detection radar site, and it was connected directly with the North American Air Defense Command, or NORAD. McManus and everyone else who worked here knew the remote, isolated facility as “J-Site.”

“Our job at J-Site amounted to ninety percent boredom and ten percent panic,” Gene McManus recalls. “The panic was if the power went off or if there was a missile scare.”

J-Site was part of ARPA’s secretive 474L System Program Office, which was responsible for developing techniques and equipment to track all objects in space and any ICBMs that might be coming in
over the North Pole. The Air Force ran the place, and McManus technically worked for RCA, Radio Corporation of America, under its defense contractor division, RCA Service Company.

The Arctic environment played a role in everything, McManus explains. J-Site was located thirteen miles from the main Defense Department base in Thule, Greenland, an area that was landlocked by ice nine months of the year. For roughly four of those months, the sun never came up over the horizon and the temperature stayed around -40 degrees Fahrenheit. There was darkness all day and all night, the black sky interrupted only by the low-rising moon. For the two hundred people who worked at J-Site each day, the commute was called “the coldest thirteen miles on wheels.”

The J-site workers—mostly radar technicians and maintenance crews—rode to the BMEWS site in a twelve-bus convoy that always traveled in tight formation. If any bus were to fall behind, get stuck, or have engine failure, it would not take long for the passengers to freeze to death. In a phase-one blizzard, which was common, bus drivers battled 70-mile-per-hour winds and maintained visibility of about fifty feet. But if a phase-three blizzard hit, the worst kind of storm, with winds up to 120 miles per hour, visibility was reduced to inches, and the road turned into a giant snowdrift. Bus drivers had to slow down to a treacherous 10-mile-per-hour crawl. Driving slower meant the bus engine could stall. Driving faster meant the bus driver might drive off the road into deep snow. One Christmas, Gene McManus and his fellow crewmembers got caught in a phase-three blizzard, and the commute that normally took thirty to forty minutes took thirteen hours. “The anemometer [wind meter] at the BMEWS site pegged at a hundred and sixty-five miles per hour,” McManus recalls. He and his crew got stranded at J-Site, which was particularly unfortunate because the Bob Hope USO tour was visiting Thule Air Base that holiday, and instead of seeing the show live, the stranded J-Site workers had to listen to the gala over the public address system.

J-Site was a futuristic-looking environment with some of the most modern, most powerful technical equipment in the world, perched high on a frozen, treeless bluff overlooking the Wolstenholme Fjord. Four massive radar antennas, each 165 feet high and 400 hundred feet long, were programmed to track objects three thousand miles out. When McManus arrived in the spring of 1961, workers were building the radome, a bright white 150-foot-tall microwave radar dome that looked like a giant golf ball made of honeycomb pieces in the shape of pentagons and hexagons bolted together.

In the summer it was beautiful. “We would watch the icebergs calve from the glaciers, and when the fjord thawed, the water was clear blue,” McManus remembers. He and other technicians would take summer walks around J-Site. The landscape was barren above the Arctic Circle, but when the snow melted, from June to early September the tundra bloomed with moss, cotton, and poppies. Sometimes you could see arctic foxes and hares if you had sharp eyes.

At J-Site there were nine buildings attached by enclosed roadways, like tunnels. Because the ground was permanently frozen, nothing was built underground. J-Site was a self-supporting facility with its own mess hall, receiving docks, and machine shop—all in support of the computer rooms, which were the heart of the BMEWS facility. The outpost required 85 megawatts of electricity “to provide full power to the radar and auxiliary equipment, lights, and computers,” McManus explains, enough wattage to power about fifteen thousand U.S. homes. For this, J-Site had its own power source in the oil-fired turbines on a Navy ship at anchor in the bay. “The heat generated by the power ship kept the water in the ship’s permanent mooring thawed, even at minus forty degrees,” says McManus.

It was Gene McManus’s job as an electronics technician to take care of the cables at J-Site, and these were far from any old cables. “Hundreds of miles of inch-thick multi-conductor cable carrying control, communication, and radar receiver information [were]
laid perfectly straight in the cable tray, never crossing over or under another,” McManus recalls. “Each was tied down at precise intervals, with the knots in the cable ties all facing the same direction. When the cables had to bend around corners, the radius of the bends of all the cables in the tray were exactly the same.” Precision was everything. The information flowing through these cables could start or prevent World War III.

“The ten percent panic part of the job came when something unusual was happening with the electricity,” says McManus. “Once we had a water leak in one of the antennae, in one of the waveguides. The power was down for about fifteen minutes.” It was nerve-racking, but it was nothing compared to what happened the third day J-Site went into twenty-four-hour operational mode, on October 5, 1960.

Three thousand miles from J-Site, deep inside Cheyenne Mountain in Colorado Springs, a clock on the wall read 3:15 p.m. Air Force colonel Robert L. Gould was sitting in the NORAD War Room when an alarm light flashed red. NORAD, or North American Air Defense Command, was an organization created in 1958 by the United States and Canada to defend against a Soviet attack. The War Room was where military personnel monitored airspace for ICBMs and incoming Soviet military aircraft. Colonel Gould was facing a freestanding, twelve-by-twelve-foot transparent plastic display board with a map of North America and Eurasia drawn on it. Above the map was an alarm-level indicator made up of five red lights. Nearby, Air Force technicians monitored information coming in from the BMEWS J-Site at Thule.

Suddenly, the Level Three light flashed. Had the Level One light flashed, NORAD protocol would have required Colonel Gould to “assemble the battle staff [and] watch closely.” If the light had flashed on Level Two, Gould would know “the contact is significant. Be ready to move in seconds.” Instead, the alarm system sounded at Level Three, which required Gould immediately to
contact the Joint Chiefs of Staff in Washington, the Chiefs of Staff Committee in Ottawa, and Strategic Air Command (SAC) headquarters in Omaha, Nebraska. A flashing Level Four was something every individual in the War Room knew about from training but dreaded ever having to deal with because it meant “You are apparently under attack.” A Level Four flashing light required an officer to “bring defense weaponry up, warn SAC to prepare its ICBMs for launching, get its bombers off the ground and turn loose the airborne alert force.” Level Five was the endgame. It indicated “it is 99.9 percent certain that you are under ICBM attack.”

With Level Three flashing, Colonel Gould picked up the telephone in the War Room. As he waited to connect with NORAD’s commander in chief, Air Force general Laurence Kuter, the alarm level suddenly went to Level Four, then Level Five. Gould quickly learned that General Kuter was flying over South Dakota and could not be reached, so he was instead put in touch with NORAD’s deputy commander, Air Marshal Charles Roy Slemon, of Canada. By now, also on the line was NORAD’s chief of intelligence, Air Force brigadier general Harris B. Hull.

“Where is Khrushchev?” Air Marshal Slemon asked Brigadier General Hull.

“In New York City,” Hull quickly replied.

This changed everything. There was a moment’s pause.

“Do you have any intelligence indications that would tend to confirm the radar reports” of an ICBM attack? Slemon asked.

“None, sir,” Hull said.

What was said next remains classified.

To Air Marshal Slemon, it seemed extremely unlikely that the Soviets would strike North America when Premier Khrushchev was in New York City, at the United Nations. But Slemon also believed that an attack could not be ruled out entirely and that it was time to get the BMEWS J-Site on the phone.

The technicians at J-Site, who were manning the IBM 7094
computers that received data from the radar, analyzed it, and made calculations were seeing very strange radar returns. A radar echo from an incoming ICBM took one-eighth of a second to receive. These radar returns were seventy-five seconds long. How could anything be that far away? But whatever it was that was coming over the horizon, according to the computers there were literally thousands of them. Here at J-Site, where environment was everything, someone thought of looking outside. There, coming up over the horizon, over Norway, was a huge rising moon.

The BMEWS had not malfunctioned. It was “simply more powerful than anyone had dreamed,” said a NORAD spokesman after the story broke on December 7, 1960. The “BMEWS—thought to have a range up to three thousand miles—had spotted the moon nearly a quarter of a million miles distant,” explained reporter John Hubbell. The J-Site computers had not been programmed to read or express that kind of distance and instead “divided three thousand miles into the precise distance to the moon and reported the distance left over—twenty-two hundred miles—as range.”

It was a defining moment in the history of weapons development and the future of man and machine. A computer had reported that a thousand-strong Soviet ICBM attack was under way. And a human, in this case Air Marshal Charles Roy Slemon, used his judgment to intervene and to overrule. At J-Site, the ARPA 474L System Program Office worked with technicians to teach the BMEWS computers to reject echoes from the moon.

On October 5, 1960, nuclear Armageddon was averted, but the underlying reality of national defense was that the scientists who had created the hydrogen bomb had created a weapon against which there was no defense. In ARPA’s first years as an agency, its single biggest program was Defender, with a mission to advance antiballistic missile technology and further develop “early warning
systems” like the one at J-Site in Thule. Defender began with a publicly announced first-year budget of $100 million, roughly half of ARPA’s entire budget. This figure was misleading, as Herb York explained in now declassified memos, because it included only research and development costs, not operational costs. In its first two years alone, the Pentagon spent closer to $900 million on Defender, York said, roughly $7.3 billion in 2015.

The Defender program, also called Ballistic Missile Defense (BMD), was ARPA’s most important national security program and the one that received the most press. People wanted to believe that the brilliant scientists who had created weapons of mass destruction in the first place could create a means to defend against them, especially now that the Soviets had an arsenal of their own. To Herb York, the situation was dire, mostly because of the time frame involved. York ordered ARPA scientists to determine the
exact
amount of time it would take for a Soviet ICBM carrying a megaton warhead to travel from a launch pad in Russia to a target in Washington, D.C.

In a secret dossier, “Assessment of Ballistic Missile Defense Program” (PPD 61-33), obtained through the Freedom of Information Act (FOIA) and not known to have been reported before, ARPA mathematicians whittled that number down to an exact figure—a mere 1,600 seconds. It seemed impossibly fast. Just twenty-six minutes and forty seconds from launch to annihilation.

The ARPA report chronicled the journey of the nuclear-armed ICBM in its three stages: boost, midcourse, and terminal phase. The initial boost stage took three hundred seconds—five minutes. This included the time it took for the rocket to fire up off the launch pad, head skyward, and reach cruising altitude. The second stage, called midcourse, lasted 1,200 seconds—twenty minutes. This stage included the time it took for the missile to travel in an arc-like trajectory over the planet at an altitude of approximately eight hundred miles above sea level. The final stage was called the
terminal stage. It accounted for the last one hundred seconds of flight—1.6 minutes. This terminal phase began when the warhead reentered the earth’s atmosphere and ended when it struck its target—an American city. Sixteen hundred seconds. That was it.

The secret “Assessment of Ballistic Missile Defense” came with a stern forewarning: “The nuclear-armed ICBM threatens us with annihilation; the stakes are so high that we must explore every alternative of strengthening our military posture.” One of the great tragedies, or ironies, here was that defending against a single ICBM was actually not too difficult a task. According to the authors of the ARPA report, the Army’s antiballistic missile system, called Nike-Zeus, gave “high confidence that… targets [i.e., incoming Soviet ICBMs] could be destroyed.” The problem was numerical, the scientists said. It was the sheer volume of megaton weapons in existence—with more still being engineered—that made the situation so hopeless. “The most important limitation of [Nike-Zeus] is that its firepower will probably not be able to handle the number of simultaneous targets which can reasonably be expected in an all-out war with the USSR,” the scientists wrote.

For Herb York, it was time to go back to the Supermen of hard science. Several of the men from Project 137 had formed a defense consulting group of their own. They called themselves the Jason scientists.

“I suppose you could say I started Jason,” said Murph Goldberger in a 2013 interview, at the age of ninety-one. The former Manhattan Project member, former science advisor to President Johnson, former president of the Federation of American Scientists, and the first American scientist to travel to communist China on an official government-sponsored science mission, among other impressive feats, was living with his full faculties intact at a retirement home in La Jolla, California, called Casa de Mañana, or House of Tomorrow. Eating Hungarian goulash in a dining room
filled with people of a similar age, and with a commanding view of the vast Pacific, Goldberger explained, “I was Jason’s first director or president. It was an impressive group. We were scientists committed to solving defense problems.”